[unreadable] The wide variation in the maintenance of body fat has a complex network of causes spanning cellular metabolism, insulin signaling, endocrine regulation, and behavior. This project aims to apply a systems biology approach to model variation in fat homeostasis in Drosophila melanogaster by threading together data at these multiple levels. We make use of the genome sequence information to design accurate assays to quantify gene expression levels in 90 candidate genes in relevant pathways. These measurements will be made in a set of 282 lines of globally distributed origin to quantify the genetic basis for natural variation in lipid storage by relating it to variation in transcript abundance and activities of enzymes in relevant metabolic pathways. Additional metabolic and endocrine phenotypes that will be measured include body mass and composition, lipid profiles, ecdysteroid levels, metabolic rate, and work capacity, quantified by maximal free flight power output. In the first aim, metabolic and physiological models will be fitted by Bayesian methods to identify associations among these variables, with emphasis on relating variation in gene expression levels to the fat storage phenotypes. The second aim is to determine the evolutionary conservation of the fat homeostasis regulatory network by quantifying gene expression and variation in fat storage phenotypes in lines selected for starvation resistance, in a set of D. simulans crosses, and in the 10 additional Drosophila species whose genomes are being sequenced. Applying similar modeling efforts to those performed in the first aim will provide an understanding of the underlying causes for evolutionary stability of fat regulation. As a third aim, genetic and nutritional perturbations will be applied to further test the robustness of the genetic pathways underlying lipid storage. We will perform a screen to find P-element derived deficiencies that impact lipid storage in flies and the FLP-FRT method will be used to generate targeted deficiencies of 10-15 key genes. All these mutants will be scored for metabolic phenotypes to determine how these genetic lesions interact with the known candidate genes. Metabolic phenotypes will be measured in flies reared on four dietary regimes. Through our modeling efforts, these data will provide an integrated, quantitative picture of how variation in lipid storage is mediated by variation in expression of the underlying genes. We anticipate that these results will have relevance to public health by identifying which particular genes and pathways are most relevant to variation in fat storage homeostasis, and by developing a systems biology approach to make inferences from data that span cellular, endocrine, and whole-organism attributes [unreadable] [unreadable]